Adi Haviv

This talk addresses the challenge of quantifying originality in text-to-image (T2I) generative diffusion models, with a focus on copyright originality. We propose a method that leverages textual inversion to measure the originality of an image based on the number of tokens required for its reconstruction by the model. Our approach is inspired by legal definitions of originality and aims to assess whether a model can produce original content without relying on specific prompts or having the training data of the model. We demonstrate our method using both a pre-trained stable diffusion model and a synthetic dataset, showing a correlation between the number of tokens and image originality. Further, we delve into the ability of models to generate unseen content and assess their generalization capabilities. Our findings suggest that T2I models are capable of generating original content and highlight the importance of training models on diverse datasets. This work contributes to the understanding of originality in generative models and has implications for copyright infringement cases.